For some time, fecal occult blood (FOB) testing has been incorporated into a large number of colorectal cancer screening programs in the United States and internationally. In these programs, the guaiac-impregnated paper slide is the most commonly used FOB test used to detect occult stool blood. Distribution of the FOB test in colorectal cancer screening programs is predicated on the test's capacity to indicate the presence of early-stage tumors. FOB test performance in screening program target populations, however, has been disappointingly low. The proposed study is a controlled investigation of participation in a colorectal cancer screening program that is offered routinely by an individual practice association (IPA) HMO to its member population of men and women 50 or more years of age. A telephone survey of 500 men and women randomly selected from the screening program target population will be conducted initially to gather data relevant to FOB test performance. These data will include sociodemographic characteristics; health knowledge, attitudes and behaviors; past health behavior and future behavioral intent; and perceived social support. FOB test performance in two different screening program models - one based in a centralized screening office and one based in physician offices - will be analyzed. The effect of mailed and telephone reminders also will be studied. Physician support and reminders are studied in controls (n=200) and study group (n=200). Studies of participation relative to colorectal cancer screening have suffered from a variety of weaknesses, including: (1) age variance between study populations; (2) the analysis of different health behaviors and explanatory variables from study to study; and (3) failure to employ multivariate statistical techniques. As a result, few clear conclusions have been drawn regarding factors that influence participation in colorectal cancer screening programs. In the proposed study, these concerns are addressed by defining the study population by age (men and women 50 or more years of age), by specifying the behavior to be studied (FOB test performance), and by using multivariate statistical techniques in data analysis. Conclusions drawn from this study will be used to develop interventions to increase FOB test performance in populations offered occult blood testing as part of cancer screening programs.